Authorship-Attribution

Various algorithms for text classification

Performing an authorship attribution task on Donald Trump’s tweets

Trump’s tweets are not explicitly labeled It is known, however, that Trump was using an android phone while the staffers were most likely to use an iPhone. Luckily, the device information is part of the data available via the Twitter API, hence the device used can be used as an authorship label.

In this task using a number of supervised machine learning classifiers in order to validate the hypothesis about Trump tweeting habits.

The attached document explains the task process:

• Data Preprocessing and Exploration
• Features Extraction
• Methods
• Experiments

Algorithms:

• Logistic Regression
• Support Vector Machine algorithms - Linear and nonlinear kernels
• Random Forest
• Feed Forward Neural Network in Pytorch